NPU Deep Learning Compiler Engineer

Who we are

We are a team that designs architectures and develops software for high-performance AI semiconductors (NPUs) used in autonomous driving and robotics.

We focus on architecture and Instruction Set Architecture (ISA) design, as well as deep learning compiler and performance simulator development, while driving architectural optimization through Design Space Exploration (DSE).

Responsibilities

This role is responsible for developing graph-level optimizations and lowering passes for deep learning models, as well as building runtime environments for efficient execution of compiled workloads on NPUs.

In this role, you will be responsible for designing interfaces between frontend and backend compilers, and collaborating with NPU architects and backend compiler engineers to optimize end-to-end performance.

■ Frontend Compiler

• Developing graph-level optimization pass
• Developing operator-level lowering pass
• Optimizing dataflow and memory hierarchy scheduling
• Designing and enhancing intermediate representation (IR) and compilation pipelines

■ Runtime

• Developing runtimefor executing compiled code
• Developing memory and resource management features
• Supporting performance profiling and bottleneck analysis from a runtime perspective

■ Code Generation

• Developing interfaces for backend compilers
• Improving end-to-end performance and quality
• Analyze bottlenecks and identify improvement areas at frontend/backend boundaries
• Optimize end-to-end performance through collaboration with backend, runtime, and hardware architects

Minimum Qualifications

• Master's degree in Artificial Intelligence, Electrical Engineering, Computer Science, or a related technical field
• Minimum 2 years of experience in deep learning compiler (graph-level) optimization, lowering pass development, runtime environment development (a relevant degree may be considered equivalent to professional experience)
• Experience in one or more of the following areas: Graph/IR transformation, runtime, or inference performance optimization
• Proficiency in C/ C++ and Python

Preferred Qualifications

• Ph.D. in compiler development targeting NPUs or GPUs
• Experience in fabless semiconductor companies or related organizations specializing in AI accelerators (e.g., NPUs, GPUs)
• Experience extending and customizing open-source compiler stacks (e.g., TVM, MLIR, LLVM, XLA) for internal use
• Experience developing compilers or runtimes for deep learning accelerators
• Understanding of state-of-the-art AI model architectures and low-precision inference
• Experience designing domain-specific languages (DSLs) optimized for AI accelerators or specific domains and integrating them into compiler pipelines
• Proficiency in English (fluent in both written and oral business communication)

Others

[Notes]

• In accordance with the employment protection laws, eligible candidates (disabled individuals, veterans, etc.) will receive preferential benefits during the process.
• Applications are accepted only through the Hyundai Motor Company career page, and any other methods are not allowed.
• Submitting your applications early prior to the deadline is recommended to prevent mass traffic and submission issues.
• Additional assignments may be given during the interview rounds, if necessary.
• The recruitment schedule may be adjusted depending on the progress of the recruitment process.
• The expected position level specified in the job description may change based on each individual's experience and evaluation results.
• For any inquiries on recruitment, please refer to the [FAQs] or submit a ticket through '1:1 Inquiries' on our webpage.

[Precautions]

• If any of the following conditions apply, your acceptance may be revoked, or you may be subject to disadvantages in the selection process: 
• If the information submitted during the recruitment process (including the application form) differs or cannot be proven in documentation
• If you are ineligible to work under company regulations
• If you are ineligible to travel abroad (for men, Korean nationals who have not completed military service or are not exempt from military service by the joining date)
• If you are unable to join the company on the start date designated by the company upon receiving the final offer
• Our company respects the trade secrets of third parties and strives to comply with relevant laws and regulations, such as The Unfair Competition Prevention and Trade Secret Protection Act. Please be careful not to disclose any confidential information related to your former/current workplace or to violate the confidentiality obligation of your former/current workplace, during the recruitment process. Violation may result in disadvantage in the recruitment process.

Position: Research Engineer, Senior Research Engineer